3D Object Detection on nuScenes v1.1 (val)
59.2NDSSparseBEV
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| SparseBEVBackbone=ResNet101, Input Size=1408 x 512, Epochs=24, Perspective pretraining=true2023.08 | 59.2 | 50.1 | 0.562 | 0.265 | 0.321 | 0.243 | 0.195 | |
| SOLOFusionBackbone=ResNet101, Input Size=1408 x 512, Epochs=90, CBGS=true2023.08 | 58.2 | 48.3 | 0.503 | 0.264 | 0.381 | 0.246 | 0.207 | |
| SparseBEVBackbone=ResNet50, Input Size=704 x 256, Epochs=36, Perspective pretraining=true2023.08 | 55.8 | 44.8 | 0.581 | 0.271 | 0.373 | 0.247 | 0.19 | |
| StreamPETRBackbone=ResNet50, Input Size=704 x 256, Epochs=60, Perspective pretraining=true2023.08 | 55 | 45 | 0.613 | 0.267 | 0.413 | 0.265 | 0.196 | |
| Sparse4DBackbone=ResNet101-DCN, Input Size=1600 x 900, Epochs=48, Perspective pretraining=true2023.08 | 55 | 44.4 | 0.603 | 0.276 | 0.36 | 0.309 | 0.178 | |
| SparseBEVBackbone=ResNet50, Input Size=704 x 256, Epochs=362023.08 | 54.5 | 43.2 | 0.606 | 0.274 | 0.387 | 0.251 | 0.186 | |
| Sparse4Dv2Backbone=ResNet50, Input Size=704 x 256, Epochs=1002023.08 | 53.9 | 43.9 | 0.598 | 0.27 | 0.475 | 0.282 | 0.179 | |
| BEVDepthBackbone=ResNet101, Input Size=1408 x 512, Epochs=90, CBGS=true2023.08 | 53.5 | 41.2 | 0.565 | 0.266 | 0.358 | 0.331 | 0.19 | |
| SOLOFusionBackbone=ResNet50, Input Size=704 x 256, Epochs=90, CBGS=true2023.08 | 53.4 | 42.7 | 0.567 | 0.274 | 0.511 | 0.252 | 0.181 | |
| BEVPoolv2Backbone=ResNet50, Input Size=704 x 256, Epochs=90, CBGS=true2023.08 | 52.6 | 40.6 | 0.572 | 0.275 | 0.463 | 0.275 | 0.188 | |
| BEVFormerBackbone=ResNet101-DCN, Input Size=1600 x 900, Epochs=24, Perspective pretraining=true2023.08 | 51.7 | 41.6 | 0.673 | 0.274 | 0.372 | 0.394 | 0.198 | |
| BEVStereoBackbone=ResNet50, Input Size=704 x 256, Epochs=90, CBGS=true2023.08 | 50 | 37.2 | 0.598 | 0.27 | 0.438 | 0.367 | 0.19 | |
| PETRv2Backbone=ResNet50, Input Size=704 x 256, Epochs=602023.08 | 45.6 | 34.9 | 0.7 | 0.275 | 0.58 | 0.437 | 0.187 | |
| DETR3DBackbone=ResNet101-DCN, Input Size=1600 x 900, Epochs=24, Perspective pretraining=true2023.08 | 43.4 | 34.9 | 0.716 | 0.268 | 0.379 | 0.842 | 0.2 |